Here's my suggested test protocol again:
1. Claimant is given one of two cards at random by Observer A and "sends" it telepathically to receiver who is connected to polygraph machine. Card is recorded by observer A.
2. At the same time receiver is shown both cards in random order by Observer B (out of sight of claimant) and asked if that is the one he is receiving, and answers "no" to both. The order in which the two cards were shown is recorded by Observer B.
3. Repeat, say, 20 times. So Observer A has list of 20 cards, Observer B has list of 20 pairs of cards.
4. Claimant is shown the polygraph readout on which the time at which each pair of answers was made has been marked and identifies which answer of each pair he thinks indicates the receiver was telling the truth, and which is the lie in each case. Observer B listens and strikes out either the first or second card from his list, according to which the claimant says is the lie.
5. We now have two lists which should be the same if the claimant has successfully sent the identity of each card telepathically to the receiver.
It's probably not perfect but unless I'm missing something (which is entirely possible) it should be acceptable to JREF.
This is a cat ship test in a subtly different way except that there is a third party operating the polygraph. Why would the tea leaf reader have someone else reading the tea leaves or why would I have someone else reading the poly?
It also misses the point of poly testing as I have stated before – the poly works on actions you have done rather than on your state of mind. It detects “Did you rob the bank” not “Did you rob the bank” when you did not but did think about it.
2. At the same time receiver is shown both cards in random order by Observer B (out of sight of claimant) and asked if that is the one he is receiving, and answers "no" to both
This tests their state of mind. They would have to perform an action for the poly to work at its best. The cat ship test works this way by each person (TX and RX) has to write down the word they have randomly chosen.
Thanks for the info Pixel42 but I fail to see the problem with my cat ship test below. I do however like some of your insight. Others just criticise from what seems like a “no understanding” point of view.
Two words are chosen (cat and ship) and written down on folded post it notes so that the words cannot be seen.
The TX (transmitter) and RX (receiver) each have a choice to pick one of the two words at random as they both have two post it notes, each with cat or ship written on the inside.
After a single random choice has been made by the TX and RX, their chosen word is written down by themselves on a piece of paper.
A GSR or poly is connected to the RX and then two questions are asked by the TX.
Q1) Did you write down the same word as I wrote down?
The RX should answer "No"
The GSR response is noted.
Q2) Did you write down a different word to the word I wrote down?
The RX should answer "No"
The GSR response is noted.
The higher GSR response indicates which word that the RX has - high reading on question Q1, they have the same word, high reading on Q2 they have the other word that was written down by the TX.
If the TX had written down "Ship" and the RX had a high reading on Q2 then the RX has "Cat".
The TX who would also interpret the GSR or poly to determine which word the RX had written down.
The test would be conducted a number of times, say 10 in this example. The result would be the TX has a list of his randomly chosen words. Next to each word would be a word that the TX calculated that the RX has. The RX would have a list of their randomly chosen words. The lists would be compared to see how may time the TX predicted word list matched the RX list and then accuracy of the test calculated.
For instance in my test with the Doctor, I would have written ship (my randomly chosen word) and cat (the word I predicted that the Doctor has chosen). The Doctor would have written down cat. In the above test this list would be ten lines long. My predicted word would then be compared to the Doctors written word.
In each test the cat and ship words should be changed so as to avoid confusion from one test to the other. i.e. second test would be tree and rock etc.
In this case you would not expect a 100% hit rate, but still significantly better than the 50% chance hit rate
So you accept that the GSR has a better than 50% hit rate and therefore my test with the Dr had some validity and your statement “Either way it's clear that golfy doesn't have any telepathic ability” was actually wrong. GSR accuracy (say for arguments sake 70%) x 1 test is 70% chance the Dr could hear me.
I have done some test with a friend today on the GSR with a heads tails test with me saying whether I had heads or tails verbally and him writing down his choice of heads or tails with me not being able to see his choice. There was £5 at stake if he could beat the GSR out of 5 attempts.
Initially the first reading was similar to the Doctor, no indication on one and full scale on the other. My prediction from looking at the GSR produce the correct result.
The second test was similar, zero reading and about 15 of the 18 LEDs, again correct prediction.
The third test was around 8 LEDs and zero reading, again correct prediction.
The fourth test was 3 LEDs and zero reading, again correct prediction.
The fifth test was zero on both questions.
This perhaps shows the inadequacy of repeating the same test again and again. Heads tails once, then cat ship, then tree rock etc may have given better results.
It does also indicate that the GSR is possibly very accurate to start with but its effectiveness falls away as the RX gets used to the testing. Overall the GSR was 80% accurate with very limited testing, so not a particularly reliable figure, but interesting.
Also I did not but will do next time, explore what happens as the RX start to beat the GSR after a number of tests if that is waht happens. Does their excitement as they start to beat the GSR then start reintroducing stress due to their will to gain the £5 after they have got used to the test, which enabled them to stay calm whilst lying. As their expectation rises as they start to beat the test and get closer to winning the £5, it may then be harder to calmly lie with no GSR indication as they get to a winning position as this adds pressure into the testing.
Also perhaps the addition of a small electric shock every time the correct prediction is made (from a toy shocker, nothing dangerous) would also keep the RX in a state of anxiety to not be cault "lying" on each answer, adding reliability to the test. The test results may fade because there is no punishment involved in the test as there would be in a poly test if you were a bank robber.
15 tests instead of 5 may have been more interesting to see what the trend does. Also this is not the same stress level as only £5 was involved as a person who is deliberately lying about hearing me such as my Doctor or others may be doing. As they are in a position of authority, the consequences of being proven to be a liar would be fare more dire than a £5 stake which may keep the effectiveness of the test at a high level question after question.
golfy
